UC Berkeley Phonlab Annual Report
Total Page:16
File Type:pdf, Size:1020Kb
UC Berkeley UC Berkeley PhonLab Annual Report Title Phonetics vs. Phonology in Loanword Adaptation: Revisiting the Role of the Bilingual Permalink https://escholarship.org/uc/item/83f5j51f Journal UC Berkeley PhonLab Annual Report, 4(4) ISSN 2768-5047 Author Chang, Charles B Publication Date 2008 DOI 10.5070/P783f5j51f eScholarship.org Powered by the California Digital Library University of California UC Berkeley Phonology Lab Annual Report (2008) Phonetics vs. Phonology in Loanword Adaptation: Revisiting the Role of the Bilingual∗ CHARLES B. CHANG University of California, Berkeley 0. Introduction In recent studies of loanword adaptation, two main sides have emerged. On the one hand, phonetic accounts emphasize the influence of low-level perceptual factors in the mapping of source language (L2) forms to receptor language (L1) forms. On the other hand, phonological accounts contend that foreign words are incorporated into a language on the basis of phonological similarity between L1 and L2 phonemic categories by bilinguals with access to the phonology of both L1 and L2. In this paper, evidence from Burmese is presented in favor of an intermediate model incorporating both language-independent phonetics and language-particular phonology. The paper is organized as follows. Section 1 reviews some key studies in the ongoing debate regarding the relative influence of phonetics vs. phonology in loanword adaptation. Section 2 provides background on Burmese phonology, with special attention to phonological differences from English. Section 3 presents examples of phonological scansion and phonetic scansion in the adaptation of English loanwords in Burmese. Finally, Sections 4 and 5 discuss the results and summarize the conclusions. 1. Phonetics vs. Phonology in Loanword Adaptation Researchers endorsing a “phonetic approximation” view of loanword adaptation basically posit that perceptual (and, thereby, acoustic) similarity is responsible for the way L2 forms are mapped onto L1 forms. Though some studies may imply otherwise, phonology is not irrelevant under this view. On the contrary, it is ∗ I am indebted to Larry Hyman, Mathias Jenny, Michael Kenstowicz, Darlene LaCharité, Javier Martín-González, Lynn Nichols, Donca Steriade, Bert Vaux, Ingyin Zaw, Jie Zhang, and au- diences at SEALS XVII, BLS 34, and PLC 32 for comments, feedback, and data that have greatly improved this paper. This work was supported by a grant from the Harvard College Research Program, a Jacob K. Javits Fellowship, and a National Science Foundation Graduate Research Fellowship. Any remaining errors are mine. 43 UC Berkeley Phonology Lab Annual Report (2008) precisely the perceptual biases from the L1 phonological system that are generally thought to result in unfaithful perception of L2 forms. In a nutshell, under this view transformations occur in loanwords vis-à-vis the original L2 forms because borrowers are non-native speakers of the L2 who hear the L2 forms unreliably. In an early study representative of the phonetic approximation view, Silverman (1992) advances a two-tiered model of adaptation in which the first level involves phonetic scansion of the L2 output. He assumes that “the input to loanword phonology is merely a superficial non-linguistic acoustic signal” (1992:289), which is parsed into segments on the first level and mapped onto phonemes of the native L1 on the basis of acoustic similarity. On the second level, L1 phonological constraints are imposed upon the input, and Universal Grammar principles may apply. A notable claim of this model is that phonological knowl- edge of L2 plays no role in adaptation. Evidence from loanwords in Cantonese suggests that Cantonese speakers are unable to access the source phonological representations of incoming loanwords; thus, the role of the bilingual in loanword adaptation is said to be minimal. The strongest version of the phonetic approximation view is developed in a later series of papers by Peperkamp and her colleagues (cf. Peperkamp and Dupoux 2003, Vendelin and Peperkamp 2004, Peperkamp 2005), who argue that not just some, but all transformations in loanwords result from unfaithful L2 perception (and, thus, that these transformations are phonetic in nature). Drawing parallels between the loanword literature and the cross-linguistic speech percep- tion literature, Peperkamp argues that “loanword adaptations are basically pho- netic rather than phonological in nature, and originate in the process of phonetic decoding during speech perception” (2005:350), though she acknowledges that this hypothesis is “a strong one that might be overly simplistic” (2005:349). In contrast, the “phonological approximation” view of loanword adaptation contends that L2-to-L1 mapping occurs on the basis of phonological distance, rather than phonetic distance between categories: a foreign L2 segment is re- placed by the L1 segment that is the closest phonologically (in terms of features), which is not necessarily the segment that is the closest perceptually. The most recent proponents of this view are Paradis and LaCharité, who in series of papers (cf. Paradis and LaCharité 1997, 2008; LaCharité and Paradis 2005) argue that bilinguals, who have access to the phonology of L2, are the ones chiefly responsi- ble for introducing borrowings into a language, and that since the bilingual borrower knows the underlying representation for an L2 form, it is this phonemic representation that constitutes the input to L1. This fact lies behind much of the otherwise inexplicable data they provide on segmental adaptation. LaCharité and Paradis (2005) point out several cases of loanword adaptation where an L2 segment is replaced by the phonologically closest L1 segment instead of the phonetically closest one (independently identified in speech percep- tion and language acquisition studies). For example, English voiced stops, typi- cally realized in initial position with no voicing during closure and simply a short- lag voice onset time (VOT), are closer phonetically to Spanish voiceless stops 44 UC Berkeley Phonology Lab Annual Report (2008) (unaspirated with short-lag VOT) than to Spanish voiced stops (strongly pre- voiced with negative VOT). Therefore, if English voiced stops in loanwords were adapted phonetically, one would expect them to be mapped to Spanish voiceless stops, but on the contrary, they are mapped to the voiced stops, the phonologically closest category. The adaptation of English rhotics in Japanese and the adaptation of the English high lax vowels /ɪ, ʊ/ in a number of languages (e.g. Mexican Spanish, Parisian French) show a similar pattern. This evidence suggests that what gives rise to segmental adaptations are not relationships of phonetic similarity, but relationships of phonological similarity. The extremely low rate of deletion present in the loanword corpora examined by LaCharité and Paradis – much lower than one would predict if faulty L2 percep- tion were responsible for the changes made in loanwords – further supports the idea that borrowers’ L2 perception is, in fact, not faulty at all. 2. A Primer on Burmese Phonology 2.1. Consonant, Vowel, and Tone Inventories The segment inventories of Burmese are given in Tables 1 and 2, where strictly allophonic segments have been placed in parentheses. Table 1. Burmese consonant inventory LABIAL DENTAL ALVEOLAR PALATAL VELAR GLOTTAL PLOSIVE p pʰ b t tʰ d k kʰ g Ɂ AFFRICATE tʃ tʃʰ dʒ FRICATIVE (t)̪ θ (d)ð̪ s sʰ z ʃ h NASAL m̥ m n̥ n ɲ̊ ɲ ŋ ̊ ŋ LATERAL l ̥ l TAP/FLAP (ɾ) APPROXIMANT w̥ w j Table 2. Burmese vowel inventory FRONT CENTRAL BACK HIGH i ɪ̃ u ʊ ̃ MONOPHTHONGS MID e (ə) ɔ LOW aã DIPHTHONGS ei ẽĩ ai ãĩ au ãũ ou õũ Notable consonantal gaps in comparison to English are the lack of /f, v, ɹ, ʒ/. With respect to vowels, mid nasal vowels and /æ, ɔɪ/ do not occur. Other English vowels missing from Burmese, such as the lax vowels /ɪ, ɛ, ʊ/, have close corres- pondents in Burmese vowel allophones not included in Table 2. Burmese is a tone language, where differences between tones have to do not only with pitch, but also duration, intensity, phonation, and vowel quality (Green 2005). By most accounts (e.g. Wheatley 1987, Green 2005), there are four distinct tones: low, high, creaky, and a “checked” or glottal tone with the general features of creaky tone followed by glottal stop (cf. Table 3). The tone on schwa is neutral. 45 UC Berkeley Phonology Lab Annual Report (2008) Table 3. Burmese tone inventory TONE TRANSCRIPTION CHARACTERISTICS low à medium duration, low intensity, low/rising pitch high á long duration, high intensity, high/falling pitch, can be breathy creaky a ̰ short duration, high intensity, high/falling pitch, creaky glottal aɁ very short duration, high pitch, sharp glottal closure 2.2. Syllable Structure and Phonotactics The basic Burmese syllable structure is C1(C2)V(V)(C3), where C2 is an approx- imant. The rhyme minimally contains a monophthongal nucleus, and may also contain a diphthong. An optional coda C3 is limited to the glottal stop occurring with glottal tone. Several phonotactic restrictions apply to this basic structure. First, the glide /j/ only occurs after labials; clusters such as */tj, kj/ are ill-formed (Green 2005). Second, the diphthongs /ai, au/ only occur before coda glottal stop. Third, /ɔ/ does not occur with a glottal coda (Cornyn 1944), while lax vowel allophones [ɪ, ɛ, ʊ, ʌ] only occur with a glottal coda, or else nasalized (except [ɛ]). In addition, the configuration of a nasalized vowel followed by a coda glottal is disallowed (ibid.). Finally, a syllable with a schwa cannot stand on its own and is always bound to a following major syllable with a full vowel (ibid.). 3. Loanword Adaptation in Burmese: Phonological or Phonetic? This study focuses on a corpus of 278 adaptations comprising 193 established loanwords and 46 non-words gathered from one main Burmese-English bilingual consultant, as well as 39 adaptations cited by Wheatley (1987), Win (1998), and Green (2005).